Web Stats - How Important is accuracy?

The news that the Telegraph.co.uk site’s Unique Users shot up by 38% in a month was rightly the cause of much bragging at the Telegraph.

The fact that this coincided with a change in webstats vendor was seized upon by .. er.. The Guardian. I’d do the same in both their sets of shoes.

It highlights a little secret in the online world. It’s the most measurable of media, but that doesn’t mean the stats are accurate! ABCe will audit and confirm that you aren’t fibbing. But they don’t check to make sure you’re catching all the visits and views you could. The parallel in old media – radio – is where a radio station draws its survey area on a map and misses a few listeners on the fringes of its area. RAJAR doesn’t tell them to get their act together. It just counts the people inside the area.

Inaccurate stats don’t need to bother a website owner. The only metric that truly counts is revenue. Oh, and profit. I’ll talk about that other dirty secret another time (Facebook anyone?)

Because revenues come from either advertising or direct online sales, the raw traffic numbers aren’t terribly important. If your site is so badly constructed that users thrash around looking for something they need, you get better time-spent and more page impressions. So what? They only bought as much as from your store as on a better laid out site. With online merchants it’s just the sales that matter.

Maybe those page imps get you more advertising money? Well – not if the user is ignoring them in their ever more maddening attempts to find what they want. With advertising, the agency or client can measure what they care about – sales and traffic to their sites.

Edward Roussel, digital editor for the Telegraph, said that "It is hugely dependent on the quality of your search engine optimisation…” That’s true for my examples as well. Do you want to optimise for traffic numbers, or for revenue? There is a difference. Getting a high Google rank for a phrase that’s not going to make you money is just vanity. Optimisation should be about optimising revenue… end of.

So – well done to the Telegraph for getting a better stats package (and for using it almost correctly). Well done to the Guardian for spotting the story.
Now who’s making the most money?